Falls in Parkinson Disease: Causes, Prediction, and Prevention Ryan Duncan, PT, DPT 1 Fay Horak, PT, PhD 2 Gammon Earhart, PT, PhD 3 1 Assistant Professor,

Falls in Parkinson Disease: Causes, Prediction, and

PreventionRyan Duncan, PT, DPT 1

Fay Horak, PT, PhD 2

Gammon Earhart, PT, PhD 3

1Assistant Professor, Program in Physical Therapy, Department of Neurology, Washington University School of

Medicine

2Professor, Department of Neurology, Department of Biomedical Engineering, Department of Physiology and Pharmacology, Oregon Health & Sciences University

3Professor, Program in Physical Therapy, Department of Neurology, Department of Anatomy & Neurobiology,

Washington University School of Medicine

Program in Physical Therapy

Disclosures• Ryan Duncan, PT, DPT

• None

• Fay Horak, PT, PhD• Serves as a board member for APDM, Inc.

• Gammon Earhart, PT, PhD• None


Objectives• Describe the multi-faceted nature of postural instability

and identify key balance impairments associated with fall risk in people with Parkinson disease (PD).

• Identify risk factors related to falls and describe various outcome measures and their accuracy in predicting falls in people with PD.

• Understand the current evidence regarding the use of exercise to reduce falls and fall risk factors in people with PD.


Outline• Incidence of and risk factors for falling in PD (Duncan)

• Facets of postural instability in PD (Horak)

• Outcome measures and their properties related to accuracy of predicting falls in PD (Earhart)

• Effectiveness of exercise for reducing postural instability and falls in PD (Duncan)


Parkinson Disease (PD)• Progressive, neurodegenerative movement disorder

• Resting tremor, bradykinesia, rigidity, postural instability

• Affects approximately 1.5 million Americans

• World-wide, the number of people affected by PD expected to double between 2005 and 2030

Fahn S., Ann N Y Acad Sci, 2003Dorsey ER, Neurology, 2007


Incidence of Falls in PD• 45-68% of people with PD fall annually

• Approximately 66% of those who fall do so recurrently

• Falls expected to become “major health problem” with anticipated increase in number of individuals with PD

• Canning CG et al. Neurodegen Dis Manag. 2014. (p.204)

Latt MD, Mov Disord, 2009Paul SS, Mov Disord, 2013

Wood BH, J Neurol Neurosurg Psychiatry, 2002


Consequences of Falls in PD

Adapted from Bloem et al., 2001, Adv Neurol, 87: 209-22


Risk Factors for Falls in PD• Cognitive impairment

• Global (i.e. MMSE), Executive Function, Attention, Central Processing

• Depression• Anxiety• Prior history of falls• Balance impairment

• Static• Dynamic

• Freezing of gait / gait impairment• Physical activity• Reduced lower extremity strength / power• Difficulty with ADLs

Potential to be directly modified by physical therapists

Canning, Neurodegen Dis Manag , 2014


Insert Fay’s Slides here


Vicious Cycle

• What is the best way to predict falls in hopes of breaking this cycle?

Adapted from Bloem et al., 2001, Adv Neurol, 87: 209-22


History Repeats Itself

• Best predictor of falls = prior history of falls (Pickering RM, Mov Disord, 2007, meta-analysis)


Learning from History

Paul et al., Mov Disord, 2013


TLI: Too Little Information

• Knowing only fall history, FOG status and gait speed gives little direction for intervention

• Need for standardized outcome measures that:• Measure constructs related to falling• Are accurate in identifying those at risk and not at risk

for falls• Are feasible for clinical implementation


Old Faithful? Berg Balance Scale

• Berg Balance Scale commonly used to assess balance in people with PD• Ceiling effect• 1 faller received a perfect score• Other fallers (n=5) were in top 10% of

BBS scores

Leddy et al., Phys Ther, 2011

Overall (n=80)

Fallers (n=25)

Nonfallers (n=55)

Age 68.2 ± 9.3 68.8 ± 7.8 67.9 ± 10.0

% Male 59 64 56

H & Y Stage

2.45 ± 0.64 2.9 ± 0.71 2.3 ± 0.50


New Faithful? FGA or BESTest



New Faithful? FGA or BESTest



A Small Improvement: The MiniBESTest

Leddy et al., J Neurol Phys Ther, 2011


Prospective Fall Prediction in PD

• BESTest, Mini-BESTest, BBS, and FGA administered at baseline (n=80)

• Fall history obtained six months and 12 months from baseline assessment

Duncan et al., Parkinsons Dis, 2012














FGA

Yang et al., Phys Ther, 2014


Yang et al., Phys Ther, 2014


Gait and Falls

• Walking difficulty strongest predictor of fear of falling (Lindholm et al., BMC Neurol, 2014)

• Why not measure gait to assess fall risk?


Cage Match: MiniBESTest Vs. Gait

Duncan & Earhart, Parkinson’s Disease, 2012


Cage Match: MiniBEST vs. Gait Velocity

Duncan & Earhart, Parkinson’s Disease, 2012


Dual Task Debate

Smulders et al., J Neurol, 2012


TUG

Vance et al., Phys Ther, 2014

1-second increase in TUG = 5.4% increase in odds of reporting a fall

Nocera et al., APM&R, 2013


TUG +

Vance et al., Phys Ther, 2014


Pull Test – UPDRS


MDS-UPDRS-III

Mak & Auyeung, J Rehab Med, 2013


Mak & Auyeung, J Rehab Med, 2013


Fall-Risk Assessment

• Nothing is perfect

• Use the full picture (i.e. don’t rely solely on balance measures) • Collective interpretation of multiple balance tests has positive

results (Dibble LE Phys Ther 2008)

• Consider cognition (Paul et al. NNR 2014, Mak et al. Arch Phys Med Rehab 2014, Amar et al. Parkinsonism Relat Disord, 2014)

• Ideal world:• Twice yearly PT evaluation (Duncan et al. Parkinson’s Dis 2012)

• Patients would be tested on and off anti-PD medication (Dibble LE Parkinsonism Relat Disord 2011)



Fall Prevention in PD• Outline

• Evidence related to using exercise to reduce actual falls in PD

• Evidence related to using exercise to modify fall risk factors in PD

• Implications for physical therapy for those with PD at risk for falling


Allen NE, Mov Disord, 2011

• 2 studies included

• No effect of physical intervention on proportion of fallers compared with usual care

• One of the trials included was not designed to reduce falls


• Included 4 published full-length papers and 3 published abstracts

• Non-significant trend noted for reduction in falls reduction with PT

• No significant difference in fall reduction when comparing to PT to no intervention

Tomlinson CL, Cochrane Database Syst Rev, 2012


• RCT

• Balance training (n=28): • Self-destabilization exercise• Externally induced destabilization exercise• Postural control feedback

• Control (n=27):• Joint mobilization, motor coordination exercise, and muscle

stretching

• Both groups: 21 sessions, 3x/week, 50 minutes each session

Smania, Neurorehabil Neural Repair, 2010


• Significant differences between groups (p≤0.001):• Pre-test Post-test • Pre-test 1-Month Follow Up

Smania, Neurorehabil Neural Repair, 2010

Number of Falls Before and After Balance Training

Pre-Test Post-Test 1-Month Follow Up

Experimental 4.3 (9.1) 1.3 (4.7) 1.3 (4.7)

Control 4.6 (8.0) 4.1 (7.3) 4.1 (7.0)

* Values are mean (SD)


• RCT

• Tai Chi (n=65)• Six Tai Chi forms for first 10 weeks, then progressed to 8-form

routine

• Resistance Training (n=65)• Lower extremity strength (1-3 sets, 10-15 reps): hip, knee, ankle• Weighted vest added at week 10, 1% BW 5% BW

• Stretching (n=65)• Active control group - breathing, stretching, relaxation

• All groups exercised for 60 minutes twice weekly for 6 months

Li F, N Engl J Med, 2012


• Primary Outcome Measures• Maximum excursion (i.e. limits of stability)• Directional control (i.e. movement accuracy)

• Secondary Outcome Measures• Gait – velocity, stride length• Knee flexor/extensor strength• Functional reach test• Timed Up & Go• UPDRS III• Falls – daily fall calendars


• Incidence Rate• Tai Chi significantly different from stretching (p=0.005)• Tai Chi different from resistance (p=0.05)

Self-Reported Falls During 6 Month Intervention

Tai Chi Resistance Stretching

Any (n (%)) 19 (29) 31 (48) 26 (40)

1 3 (5) 8 (12) 4 (6)

2 4 (6) 7 (11) 2 (3)

≥ 3 12 (18) 16 (25) 20 (31)

Rate 0.22 0.51 0.62


• RCT• Intervention (n=64)

• Group strength and balance exercises 1x/weekly for 60 minutes• Home exercises twice weekly• Usual care at discretion of clinical team

• Control (n=66)• Usual care at discretion of clinical team

• Intervention period lasted for 10 weeks

• Usual care – medical/medication management, PT (i.e. exercise advice, provision of walking aids, gait training), OT (i.e. home modifications, provision of aids), or ST.

Goodwin VA, J Neurol, Neurosurg, Psychiatry, 2011


Figure adapted from: Goodwin VA, J Neurol, Neurosurg, Psychiatry, 2011


van Nimwegen, BMJ, 2013

• Multifaceted Behavioral Change Program

• ParkFIT• Activity coaching

• Education related to benefits of physical activity• Identifying and overcoming barriers to engaging in physical

activity • Systematic goal setting using health contract and logbook• Stimulation to participate in group exercises• Ambulatory monitor data with automated feedback

• ParkSAFE• Traditional PT

• Education related to benefits of PT and safety of movements• Active lifestyle not explicitly stimulated


• Falls measured as adverse events, not a primary or secondary outcome

• Participants with one or more fall:

• ParkFit: 62%

• ParkSAFE: 67%

• Take home: encouraging physical activity did not result in reduced falls, but also did not increase falls


Risk Factors for Falls in PD

• Cognitive impairment• Global (i.e. MMSE), Executive Function, Attention, Central Processing

• Depression• Anxiety• Prior history of falls• Balance impairment

• Static• Dynamic

• Freezing of gait / gait impairment• Physical activity• Reduced lower extremity strength / power• Difficulty with ADLs

Potential to be directly modified by physical therapists

Canning, Neurodegen Dis Manag, 2014


Balance Impairment

Aerobic Exercise

PT / Balance Training

Poliakoff, 2013

Smania, 2010De Goede, 2013Ellis, 2005

LE Strengthening

Allen, 2010Li, 2012


Balance Impairment

LSVT BIG ®

Ebersbach, 2010

Nordic Walking

Reuter, 2011


Balance ImpairmentOverground

Walking with and without Cues

Treadmill Walking with and without

Cues

Robotic Treadmill Walking

Allen, 2010Neiuwboer, 2007Poliakoff, 2013

Mehrholz, 2010Frazzitta, 2009

Picelli, 2013


Balance Impairment

Dance

Duncan, 2012Earhart, 2009

Tai Chi

Li, 2012


Freezing of Gait / Gait Impairment

• Multi-modal exercise program (Allen, 2010)

• Physical therapy (Tomlinson, 2012)

• Cueing• Overground gait (Nieuwboer, 2007)• Treadmill (Frazzitta, 2009)

• Dance (Duncan, 2012)


• Multi-modal exercise program• Balance exercises• Strengthening exercises• Cueing exercises (if noted to have freezing)

• 40-60 minutes, 3x/week for 6 months

Freezing of Gait

Freezing of Gait Questionnaire Scores

Pre-Test Post-Test

Exercise (n=21) 6.8 ± 5.1 5.5 ± 5.9

Control (n=24) 8.3 ± 5.8 9.4 ± 6.2

Freezing of Gait (Yes / No)

Pre-Test Post-Test

Exercise (n=21) 13 7

Control (n=24) 14 12

Allen, Mov Disord, 2010

Comfortable Gait Speed (m/sec)

Pre-Test Post-Test

Exercise (n=21) 1.07 ± 0.27 1.09 ± 0.26

Control (n=24) 1.04 ±0.25 1.06 ± 0.32


Freezing of Gait / Gait Impairment

• RESCUE trial

• Largest cueing trial to date (n=153)

• Randomized, crossover design

• 3 week home cueing program using a prototype cueing device

• Significant improvements in:

• Gait speed• Step length• Freezing of Gait severity (5.5% reduction in freezers only)

Nieuwboer, J Neurol Neurosurg Psychiatry, 2007


Freezing of Gait / Gait Impairment• PD4PD trial

• Randomized, controlled trial

• Twice weekly dance classes for one year (Argentine Tango) vs. control group

• Participants tested OFF anti-PD medication

Duncan, Neurorehabil Neural Repair, 2012


Physical Activity

• Cueing training

• RESCUE trial

• Participants wore activity monitors while training with cueing devices

• Significant improvements in time spent in static activity, dynamic activity, and walking

• Effects diminished after intervention suggesting need for continued practice

Lim, Neurorehabil Neural Repair, 2010


Physical Activity

• Multifaceted Behavioral Change Program

• No difference between groups for change in physical activity and 6 Minute Walk Test

• Significant improvement in number of hours of activity per week and Kcal expenditure per day in experimental group

van Nimwegen, BMJ, 2013


Reduced Lower Extremity Strength / Power

• Knee extensor strength

• Lower limb strength exercises (Lima, 2013; Li, 2012)• Treadmill training (Yang, 2010)• Tai Chi (Li, 2012)

• Knee flexor strength

• Lower limb strength exercises (Li, 2012)• Tai chi (Li, 2012)

• Leg extensor power

• Leg muscle power training (Paul, 2014)


• Leg muscle power training

• Leg extensors, knee flexors, hip flexors, hip abductors• 3 sets of 8 repetitions as fast as possible, supervised by PT• Twice weekly, 45 minutes each session for 12 weeks

• Control

• Low intensity exercise of trunk, leg flexors and extensors, and hip abductors at home

• 2 sets of each exercise twice weekly, started at 8 reps per set• Increased reps by 2 every 4 weeks

Paul, Clin Rehabil, 2014


Mean Change in Lower Extremity Power after 12 Weeks

Experimental Control p (effect size)

Leg Extensors (W) 55.5 ± 63.8 -2.4 ± 38.5 0.002 (1.10)

Knee Flexors (W) 29.1 ± 39.6 -1.7 ± 23.2 0.01 (0.95)

Hip Flexors (W) 75.4 ± 94.9 8.6 ± 28.4 0.007 (0.95)

Hip Abductors (W) 33.4 ± 32.1 -3.2 ± 16.5 <0.001(1.43)

• No significant changes in mobility measures (i.e. gait speed, Timed Up & Go, Freezing of Gait Questionnaire)

Paul, Clin Rehabil, 2014


Difficulty with ADLs

• A 16-month program of flexibility, balance, and functional exercises resulted in improved UPDRS II scores compared to a control group

• Cycle ergometry over 6 weeks led to significant improvement in UPDRS II scores

• Physical therapy and exercise are also helpful in improving ADL performance

Schenkman, Phys Ther, 2012Lauhoff, Disabil Rehabil, 2013Tomlinson, Cochrane Database Sys tRev, 2013


Implications for PT

• Environment

• Remove clutter, avoid narrow spaces that may cause freezing• Use nightlights• Consider prescribing an assistive device

• Be aware of medication side-effects

• Orthostatic hypotension• Hallucinations• Impulsivity

• Consider referring to neurologist, OT, SLP when cognitive deficits are present


Conclusions• Postural instability common in PD; can be measured

clinically using the BESTest and/or Mini-BESTest

• Provides system-specific approach to identification of balance impairements

• There is no such thing as a perfect fall risk assessment

• Do not rely solely on balance measures; consider cognition and other tests

• Ideally, patients with PD would receive a twice yearly PT evaluation for assessment of fall risk and change in function


Conclusions• Limited evidence suggesting that actual falls can be

reduced through participation in exercise for those with PD

• Falls are extremely complex and are not just a factor of physical mobility and performance

• Remediation of fall risk factors is possible with the use of different forms of exercise

• Suggests regular performance in an exercise program is paramount for those with PD, especially for those at risk for falls


Acknowledgements• Drs. Horak and Earhart

• Funding sources

• NIH/NINDS R01 NS077959• American Parkinson Disease Association (Greater St. Louis Chapter)• Parkinson’s Disease Foundation• Davis Phinney Foundation

• Programming Committee of the Neurology Section of the APTA


Questions?

Documents

Falls in Parkinson Disease: Causes, Prediction, and Prevention Ryan Duncan, PT, DPT 1 Fay Horak, PT, PhD 2 Gammon Earhart, PT, PhD 3 1 Assistant Professor,